What is Real User Monitoring (RUM)?

Application performance monitoring (APM) is a method for ensuring apps and websites are operating smoothly and efficiently. APM tools monitor the performance and availability of apps and alert IT teams to potential service disruptions. These tools include several monitoring solutions:

• Real user monitoring: RUM measures client-side performance experience by end users, typically measuring page load experience, interactivity, visual stability and errors across the front end.
• Synthetic monitoring: Synthetic monitoring simulates user behavior to proactively measure uptime and performance, typically for critical service endpoints.
• IoT monitoring: IoT monitoring allows IT teams to detect and fix performance problems with apps on mobile and other connected devices.
• Database monitoring: Database monitoring tracks metrics that help diagnose and address database-related performance problems.
• Infrastructure monitoring: This type of active monitoring collects data from applications’ underlying infrastructure to ensure it can fully support the apps and identify infrastructure issues before they impact the end user.
• Server monitoring: Server monitoring collects infrastructure metrics like memory usage, CPU use and throughput to understand the performance of apps and web servers.
• Network monitoring: Network monitoring provides visibility into on-premises and SaaS deployments to see how network conditions are affecting application performance.
• Cloud monitoring: This type of active monitoring collects and analyzes metrics to determine the health of cloud-based infrastructure and assets.
• Storage Monitoring: The practice of tracking the performance, availability and overall health of physical and virtual storage devices.
• IT Monitoring: Comprises a broad class of products designed to let analysts determine whether IT equipment is online and performing to expected service levels, while resolving any detected problems.

In modern DevOps environments, APM enables DevOps teams to find bottlenecks quickly and more easily troubleshoot and isolate application problems, making it an integral part of the software development process.

Digital experience monitoring (DEM), which encompasses both real user monitoring and synthetic monitoring, is important because it allows an organization to measure, improve and optimize user experience and page performance. Studies have proven that better-performing sites generally see increases in revenue, conversion and user engagement.

Most organizations understand that they must constantly strive to improve their app experience, but they struggle with execution. And waiting until there’s a problem to improve an app’s user experience often comes at a significant cost. In one study, 90% of users reported they stopped using an app because of poor performance. Rather than waiting to learn about web application issues directly from end users, end user monitoring allows organizations to proactively understand their app’s performance by monitoring user interactions and transactions with the app.

Both real user monitoring and synthetic monitoring have advantages and limitations. Synthetic monitoring’s use of behavioral scripts — rather than actual users — makes it well-suited for finding and fixing potential problems before they impact users. But the downside of synthetic monitoring is that it’s only an approximation of what a real user sees and does in the app and therefore can’t account for all real-world experiences. Real user monitoring, on the other hand, shows exactly how human users experience apps and services — but you have to wait for at least one user to encounter problems before you can address them.

Because of these differences, synthetic monitoring may be most effective for monitoring apps and websites in pre-production in order to catch problems and establish baseline performance before they go live. Once they go live, real user monitoring can then be used to provide visibility into how the websites and apps are impacting actual users.

Synthetic monitoring simulates user transactions by using behavioral scripts that emulate user flows and measure availability, functionality and performance for critical endpoints or across the entire user journey. Because this technique stages and directs an artificial user experience, it’s classified as active monitoring, whereas real user monitoring is considered passive monitoring. In practice, synthetic monitoring works like this: administrators (likely teams responsible for SLA uptime) define several checkpoints and select performance metrics. A robot client follows the predefined user journey through the app, simulating transactions that mimic human behavior, and sending information back on a page’s availability (did the URL respond?), functionality (did all the buttons work?), and performance (how fast were page resources loaded?). Typically, teams set up alerts to notify them of outages for critical service endpoints, which can trigger their incident response.

The information that synthetic monitoring collects can help answer a variety of questions, from basic uptime (is the website up?) to performance (how fast is the site right now?) and even ongoing code improvement (did our recent release improve user experience?).

Synthetic monitoring has several unique advantages, helping teams to:

• Identify and address issues before they impact users: Unlike real user monitoring, which requires human user interaction, synthetic monitoring can be employed 24/7, in low traffic, and in testing before pushing code live. This proactive approach gives teams the time to recognize, troubleshoot and resolve issues before users are even aware of them.
• Create performance baselines: Synthetic monitoring can help benchmark and improve page performance by measuring critical elements of user experience, such as how fast a page loads content, interactivity, visual stability, error count and trends in performance relative to previous builds, releases or competitors.
• Monitor user transactions and business processes: With synthetic monitoring, teams can emulate and monitor the web performance of searches, form-fills, logins and other transactions from different geographies. Comparing performance data from these processes across geographies can indicate where and how to make improvements.
• Test new features: Teams can also test new app features before they go live to see how they respond under load and if any of the changes impact third-party application programming interfaces (APIs).
• Reduce mean time to repair (MTTR): MTTR represents the average time it takes to repair and restore service to functionality. The lower the MTTR for an affected service, the less likely it is to negatively impact the business. Because synthetic monitoring enables teams to get ahead of issues before they affect the user — and alerts from synthetic monitors often trigger larger incident response coordination — it’s an effective tool for reducing MTTR. 

While synthetic monitoring is a useful method for understanding how users experience a website or app, it’s impossible to simulate every  real-world scenario. Therefore, real user monitoring is essential to understanding the experience of your web application for all users in the field.

Real user monitoring collects an app’s or website’s performance metrics directly from the end user's browser via a RUM tool. Typically, the real user monitoring process works by adding a snippet of JavaScript into the header of a web page. Once you implement real user monitoring you can answer some key questions about your website or app’s user experience, such as how long pages are taking to load, or how long the server takes to send the initial HTML file. But you can also gain insights into larger trends, such as:

• Your users’ environments: People use a wide variety of devices, operating systems, and networks to access websites and apps, while expecting an optimized experience. Real user monitoring can give you a clearer picture of your users’ needs.
• How users navigate your website: Real user monitoring helps you understand the paths users may take as they travel through your website or app, allowing you to optimize the pages they are most likely to visit first and other important points in their clickstream.
• How third-party scripts are performing: Third-party scripts can provide great value by generating revenue and collecting user data, but a slow script can hinder a page’s performance or tank it completely. Real user monitoring gives you visibility into how these scripts are impacting your page performance.
• Website performance and business outcome: Time is money. With real user monitoring you can begin to measure the correlation between website speed, performance and business outcomes. While it requires the right technology and precision to quantify page performance with exact increases in revenue, websites that embed performance into their DevOps processes generally reap rewards from increased user engagement and conversion.

Real user monitoring tools gather a lot of data from a large volume of users, so they usually include data visualizations like charts and graphs to make information and insights easier to grasp. Dashboards provide intuitive visualizations of metrics such as page load times and top pages. You can also dive into individual user sessions to get to the root cause of why a particular digital journey was subpar and drill down into ways to improve it.

Following a few best practices can help you get the most from real user monitoring:

• Benchmark your website’s performance: To get your site’s or app’s performance where you want it, you have to know where it’s starting. Speed results in higher customer engagement, so get a baseline reading of your site’s load time, bounce rate and other real user metrics, so you know where you need to improve.
• Define your performance improvements: Once you have these numbers, decide what benchmarks would indicate both problems and improvement. If, for example, you notice that users are leaving when a page takes three seconds to load, you might aim to get load times down to two seconds or less.
• Tie performance back to the business: Real user monitoring should always begin with a business goal, e.g. wanting to increase your conversion rate from 3% to 5%. Then you can track your performance metrics to determine your progress toward that goal.

Real user monitoring requires the use of a third-party solution to prepare your pages to track, collect and analyze user data, then compile and produce actionable reports. There are myriad real user monitoring tools to choose from, so ask yourself these questions when considering the tools:

• What features and functions do I need? Not all real user monitoring tools offer the same features, and not all businesses have the same monitoring needs, so it’s critical to calculate upfront what you need your tools to accomplish.
  
  At a minimum, a real user monitoring tool should be able to capture the user journey, track user engagement metrics such as bounce rates, cart abandonment rates and conversion rates, as well as set and collect custom metrics uniquely important to your business.
  
  Ultimately, a real user monitoring tool should enable an organization not just to assess its performance, but to optimize it. To that end, look for a tool that can identify processes that need optimization, capture a user’s browser, device, operating system, geo-location and other data, detect performance trends and outages, and alert teams to performance incidents to help reduce MTTR.
• How easy is it to implement? Before you purchase a real user monitoring tool, make sure it will be easy to integrate with your website or app. Find out what hardware and software it supports and understand the implementation prerequisites. Find out whether or not there’s a software development kit (SDK) available for mobile apps, and understand that you may have to perform some customization if your app or website pages are particularly complex.
• Do I want unified or standalone monitoring? Using real user monitoring in conjunction with other monitoring tools such as application performance monitoring or other performance monitoring solutions provides a more holistic view of your app or website’s performance. Alternatively, you can streamline and customize your monitoring strategy by picking only the tools that make sense for your business.
• How does the tool handle data? Every real user monitoring tool collects and manages data a bit differently – all with significant implications for your monitoring strategy. Long-term data storage allows you to see historical and current trends, which provides more grist for analysis. A tool may capture and report every customer visit or only a certain number of visits. It’s important to understand the differences and how they relate to your business needs.
• How does it report and display data? Performance reports and data visualizations are critical for translating monitoring insights into action, making reporting capabilities one of the most important real user monitoring tool features. For comprehensive reports, a tool should be able to group data around multiple parameters including the user’s geo-location, network, ISP and device, among others, while also allowing you to group data based on customized metrics.
• What are the licensing terms? Whether you decide to use real user monitoring on its own or with other monitoring tools, you need to understand licensing options to get the greatest value. Some vendors offer a stand-alone real user monitoring solution, while others pair it with synthetic monitoring, or offer it as part of a suite of performance monitoring tools. Understanding your monitoring needs will help you choose the most cost-effective solution.

Your customers expect an optimal experience every time they use your service. Slow or incomplete page loads, transaction errors, and other problems aren’t just annoying, they can taint the user’s experience of your business for good. Real user monitoring can uncover the impediments to a great customer experience like no other tool, allowing you to keep your site or app at peak performance — and keep customers coming back again and again.

• Splunk Everywhere! Application Performance Monitoring
• Splunk and Synthetic Monitoring Blog
• All in on APM
• Extracting Data from Splunk Infrastructure Monitoring
• Product: Splunk Real User Monitoring
• Digital Experience Monitoring
• What is Multi-Cloud?
• What is Network Monitoring?
• An Introduction to Observability
• OpenTelemetry for Beginners
• What is Predictive Maintenance?
• Real-Time Data: An Introduction
• What is Regulatory Compliance?
• Risk Management: A Beginner's Guide